Revised International Working Group Risk Model for Survival in Polycythemia Vera
Bibliographic record
Abstract
Background In 2013, we published the first International Working Group (IWG) risk model for survival in polycythemia vera (PV), which included 3-tiered age categories, leukocytosis (≥15 x 109/L), and venous thrombosis, as risk variables (Leukemia. 2013;27:1874). In the current study, we examined the individual prognostic contribution from specific leukocyte components and mutations, in order to construct a more contemporary clinical risk model in PV and assess its interaction with genetic risk factors. Methods Study patients were retrospectively recruited from the Mayo Clinic, USA (N=1,007) and the University of Florence, Italy (N=514), in order to serve as discovery and validation cohorts, respectively. Diagnostic criteria were according to the International Consensus Classification (Blood 2022;140:1200). Mutations were screened by next-generation sequencing (NGS). Conventional statistical methods were employed (JMP Pro 17.0.0, SAS Institute, Cary, NC, USA) and optimal cut-points for age, absolute neutrophil (ANC), lymphocyte (ALC), and monocyte (AMC) counts were determined by Receiver Operating Characteristic (ROC) analysis. Akaike Information Criterion (AIC) and ROC plots were used to estimate predictive accuracy. Results Presenting features in the Mayo discovery cohort (N=1,007) included: median age 62.2 years; males 50.7%; median leukocyte count 10.8 x 109/L; leukocyte count ≥15 x 109/L 19%; palpable splenomegaly 32%; pruritus 34%; diabetes 16%; hypertension 61%; arterial thrombosis 14%; and venous thrombosis 14%. ANC, AMC, and ALC information was available in 736 patients with respective median (range) values of 7.6 (1.6-30), 0.6 (0-4.1), and 1.54 (0.19-6.9) x 109/L. At a median follow-up of 8.9 years for the Mayo discovery cohort, 372 (37%) deaths, 46 (4.6%) leukemic transformations, and 200 (20%) fibrotic progressions were recorded. Multivariable analysis (MVA) employing ROC-determined cutoff values revealed adverse survival impact for age >70 years (HR 18.3; p<0.01), age 50-70 years (HR 4.9; p<0.01), AMC ≥0.8 x 109/L (HR 1.5; p<0.01), ANC ≥8 x 109/L (HR 1.5; p<0.01), arterial thrombosis (HR 1.5; p<0.01), diabetes (HR 1.4; p=0.04), and male sex (HR 1.3; p=0.05). A similar MVA applied to the Florence validation cohort (N=514) confirmed significance for age (p<0.01), ANC (p<0.01), AMC (p=0.04), and arterial thrombosis (p<0.01), but not for male sex (p=0.8) or diabetes (p=0.15). Accordingly, age >70 years (4 points), age 50-70 years (2 points), ANC ≥8 x 109/L (1 point), AMC ≥0.8 x 109/L (1 point), and AT (1 point) were included in a revised IWG survival model using the Mayo cohort and validated by the Florence cohort: very low risk (0 points; N=67; median 37 years), low risk (1 point; N=55; median 29.2 years), intermediate risk (2-3 points; N=308; median 20.2 years), high risk (4-5 points; N=217; median 11.5 years), and very high risk (6-7 points; N=88; median 6.3 years). 20-year survival prediction performance of the revised IWG model (AIC 204; AUC 0.88) was superior to that of the 2013 IWG version (AIC 350; AUC 0.86). Mutation information at diagnosis/chronic phase was available in 270 Mayo patients: ASXL1 10%; SRSF2 4%, IDH2 2%, TP53 2%, RUNX1 1.4%. MVA of mutations identified SRSF2 (p<0.01), IDH2 (p=0.01), and ASXL1 (p=0.05), as risk factors for overall survival. The presence of any one adverse mutation was more likely in patients with AMC ≥0.8 x 109/L (24% vs. 7% incidence; p<0.01). In all-inclusive MVA, independent risk factors for OS were age >70 years (HR 14.7), age 50-70 years (HR 6.2), any one adverse mutation (HR 3.4), and ANC ≥8 x 109/L (HR 2.3), but not AMC ≥0.8 x 109/L (p=0.9), arterial thrombosis (p=0.24), or abnormal karyotype (p=0.4). Leukemia-free survival was adversely affected by SRSF2, RUNX1, TP53, and ASXL1 mutations and myelofibrosis-free survival by SRSF2 mutation. Conclusions: The currently proposed revised IWG risk model for survival in PV is globally applicable and based on four easily accessible clinical variables (Age, ANC, AMC, and arterial thrombosis). NGS-derived mutation information overshadows the prognostic relevance of monocytosis, provides additional prognostic information for overall survival, and assists in predicting leukemic and fibrotic transformations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.010 | 0.017 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.004 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.001 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.010 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".